intelligent-trading-bot/scripts/train_signals.py
244 行
11 KiB
Python
Alexandr Savinov
2022-03-20 10:09:33 +01:00
from pathlib import Path
Alexandr Savinov
2022-03-27 19:22:02 +02:00
import click
from tqdm import tqdm
Alexandr Savinov
2022-07-17 21:32:45 +02:00
import numpy as np
Alexandr Savinov
2022-03-20 10:09:33 +01:00
import pandas as pd
Alexandr Savinov
2022-03-27 19:22:02 +02:00
from sklearn.metrics import (precision_recall_curve, PrecisionRecallDisplay, RocCurveDisplay)
Alexandr Savinov
2022-03-20 10:09:33 +01:00
from sklearn.model_selection import ParameterGrid
from service.App import *
Alexandr Savinov
2022-03-27 19:22:02 +02:00
from common.classifiers import *
Alexandr Savinov
2022-07-16 11:41:27 +02:00
from common.label_generation_topbot import *
Alexandr Savinov
2022-03-20 10:09:33 +01:00
from common.signal_generation import *
"""
Alexandr Savinov
2022-03-27 19:22:02 +02:00
Input data:
Alexandr Savinov
2022-07-17 21:32:45 +02:00
This script assumes the existence of label prediction scores for a list of labels
which is computed by some other script (train predict models or (better) rolling predictions).
It also uses the real prices in order to determine if the orders are executed or not
(currently close prices but it is better to use high and low prices).
Purpose:
The script uses some signal parameters which determine whether to sell or buy based on the current
label prediction scores. It simulates trade for such signal parameters by running through
the whole data set. For each such signal parameters, it determines the trade performance
(overall profit or loss). It then does such simulations for all defined signal parameters
and finally chooses the best performing parameters. These parameters can be then used for real trades.
Alexandr Savinov
2022-03-27 19:22:02 +02:00
Notes:
Alexandr Savinov
2022-07-17 21:32:45 +02:00
- The simulation is based on some aggregation function which computes the final signal from
multiple label prediction scores. There could be different aggregation logics for example
finding average value or using pre-defined thresholds or even training some kind of model
like decision trees
- The signal (aggregation) function assumes that there two kinds of labels: positive (indicating that
the price will go up) and negative (indicating that the price will go down). The are accordingly
stored in two lists in the configuration
- Tthe script should work with both batch predictions and (better) rolling predictions by
assuming only the necessary columns for predicted label scores and trade columns (close price)
Alexandr Savinov
2022-03-20 10:09:33 +01:00
"""
class P:
in_nrows = 100_000_000
start_index = 0 # 200_000 for 1m btc
end_index = None
Alexandr Savinov
2022-03-20 10:09:33 +01:00
Alexandr Savinov
2022-07-17 21:32:45 +02:00
# Haw many best performing parameters from the grid to store
Alexandr Savinov
2022-08-13 22:38:27 +02:00
topn_to_store = 10
Alexandr Savinov
2022-03-27 19:22:02 +02:00
Alexandr Savinov
2022-03-20 10:09:33 +01:00
@click.command()
@click.option('--config_file', '-c', type=click.Path(), default='', help='Configuration file name')
def main(config_file):
Alexandr Savinov
2022-03-27 19:22:02 +02:00
"""
The goal is to find how good interval scores can be by performing grid search through
all aggregation/patience hyper-parameters which generate buy-sell signals on interval level.
Here we measure performance of trade using top-bottom scores generated using specified aggregation
parameters (which are searched through a grid). Here lables with true records are not needed.
In contrast, in another (above) function we do the same search but measure interval-score,
that is, how many intervals are true and false (either bot or bottom) by comparing with true label.
General purpose and assumptions. Load any file with two groups of point-wise prediction scores:
buy score and sell columns. The file must also have columns for trade simulation like close price.
It can be batch prediction file (one train model and one prediction result) or rolling predictions
(multiple sequential trains and predictions).
The script will convert these two buy-sell column groups to boolean buy-sell signals by using
signal generation hyper-parameters, and then apply trade simulation by computing its overall
performance. This is done for all simulation parameters from the grid. The results for all
simulation parameters and their performance are stored in the output file.
"""
Alexandr Savinov
2022-03-20 10:09:33 +01:00
load_config(config_file)
Alexandr Savinov
2022-07-17 21:32:45 +02:00
time_column = App.config["time_column"]
now = datetime.now()
Alexandr Savinov
2022-03-20 10:09:33 +01:00
symbol = App.config["symbol"]
data_path = Path(App.config["data_folder"]) / symbol
Alexandr Savinov
2022-03-20 10:09:33 +01:00
if not data_path.is_dir():
print(f"Data folder does not exist: {data_path}")
return
out_path = Path(App.config["data_folder"]) / symbol
Alexandr Savinov
2022-03-20 10:09:33 +01:00
out_path.mkdir(parents=True, exist_ok=True) # Ensure that folder exists
#
# Load data with (rolling) label point-wise predictions and signals generated
Alexandr Savinov
2022-03-20 10:09:33 +01:00
#
file_path = (data_path / App.config.get("signal_file_name")).with_suffix(".csv")
Alexandr Savinov
2022-07-17 21:32:45 +02:00
if not file_path.exists():
print(f"ERROR: Input file does not exist: {file_path}")
Alexandr Savinov
2022-03-20 10:09:33 +01:00
return
print(f"Loading signals from input file: {file_path}")
Alexandr Savinov
2022-07-17 21:32:45 +02:00
df = pd.read_csv(file_path, parse_dates=[time_column], nrows=P.in_nrows)
print(f"Signals loaded. Length: {len(df)}. Width: {len(df.columns)}")
Alexandr Savinov
2022-03-20 10:09:33 +01:00
# Limit size according to parameters start_index end_index
df = df.iloc[P.start_index:P.end_index]
df = df.reset_index(drop=True)
Alexandr Savinov
2022-03-20 10:09:33 +01:00
#
Alexandr Savinov
2022-03-27 19:22:02 +02:00
# Find maximum performance possible based on true labels only
Alexandr Savinov
2022-03-20 10:09:33 +01:00
#
Alexandr Savinov
2022-03-27 19:22:02 +02:00
# Best parameters (just to compute for known parameters)
#df['buy_signal_column'] = score_to_signal(df[bot_score_column], None, 5, 0.09)
#df['sell_signal_column'] = score_to_signal(df[top_score_column], None, 10, 0.064)
#performance_long, performance_short, long_count, short_count, long_profitable, short_profitable, longs, shorts = performance_score(df, 'sell_signal_column', 'buy_signal_column', 'close')
# TODO: Save maximum performance in output file or print it (use as a reference)
Alexandr Savinov
2022-03-20 10:09:33 +01:00
Alexandr Savinov
2022-03-27 19:22:02 +02:00
# Maximum possible on labels themselves
#performance_long, performance_short, long_count, short_count, long_profitable, short_profitable, longs, shorts = performance_score(df, 'top10_2', 'bot10_2', 'close')
Alexandr Savinov
2022-03-20 10:09:33 +01:00
months_in_simulation = (df[time_column].iloc[-1] - df[time_column].iloc[0]) / timedelta(days=30.5)
#
# Load signal train parameters
#
train_signal_model = App.config["train_signal_model"]
signal_model_grid = train_signal_model["grid"]
# Evaluate strings to produce lists
Alexandr Savinov
2023-03-11 12:20:01 +01:00
if isinstance(signal_model_grid.get("buy_signal_threshold"), str):
signal_model_grid["buy_signal_threshold"] = eval(signal_model_grid.get("buy_signal_threshold"))
if isinstance(signal_model_grid.get("buy_signal_threshold_2"), str):
signal_model_grid["buy_signal_threshold_2"] = eval(signal_model_grid.get("buy_signal_threshold_2"))
if isinstance(signal_model_grid.get("sell_signal_threshold"), str):
signal_model_grid["sell_signal_threshold"] = eval(signal_model_grid.get("sell_signal_threshold"))
if isinstance(signal_model_grid.get("sell_signal_threshold_2"), str):
signal_model_grid["sell_signal_threshold_2"] = eval(signal_model_grid.get("sell_signal_threshold_2"))
# Disable sell parameters in grid search - they will be set from the buy parameters
if train_signal_model.get("buy_sell_equal"):
signal_model_grid["sell_signal_threshold"] = [None]
signal_model_grid["sell_signal_threshold_2"] = [None]
Alexandr Savinov
2022-03-27 19:22:02 +02:00
performances = list()
for signal_model in tqdm(ParameterGrid([signal_model_grid]), desc="MODELS"):
Alexandr Savinov
2022-03-27 19:22:02 +02:00
#
# If equal parameters, then derive the sell parameter from the buy parameter
Alexandr Savinov
2022-03-27 19:22:02 +02:00
#
if train_signal_model.get("buy_sell_equal"):
signal_model["sell_signal_threshold"] = -signal_model["buy_signal_threshold"]
#signal_model["sell_slope_threshold"] = -signal_model["buy_slope_threshold"]
signal_model["sell_signal_threshold_2"] = -signal_model["buy_signal_threshold_2"]
Alexandr Savinov
2022-03-20 10:09:33 +01:00
signal_model["rule_type"] = App.config["signal_model"]["rule_type"]
#
# Do not aggregate but assume that we have already the aggregation results in the data
#
trade_score_column_names = []
sa_sets = ['score_aggregation', 'score_aggregation_2']
for i, score_aggregation_set in enumerate(sa_sets):
score_aggregation = App.config.get(score_aggregation_set)
if not score_aggregation:
continue
trade_score_column = score_aggregation.get("trade_score")
trade_score_column_names.append(trade_score_column)
#
# Apply signal rule and generate binary buy_signal_column/sell_signal_column
#
if signal_model.get('rule_type') == 'two_dim_rule':
Alexandr Savinov
2023-03-11 17:43:59 +01:00
apply_rule_with_score_thresholds_2(df, signal_model, trade_score_column_names)
else: # Default one dim rule
Alexandr Savinov
2023-03-11 17:43:59 +01:00
apply_rule_with_score_thresholds(df, signal_model, trade_score_column_names)
Alexandr Savinov
2022-06-26 17:32:41 +02:00
Alexandr Savinov
2022-03-27 19:22:02 +02:00
#
# Simulate trade and compute performance using close price and two boolean signals
Alexandr Savinov
2022-03-27 19:22:02 +02:00
# Add a pair of two dicts: performance dict and model parameters dict
#
performance, long_performance, short_performance = \
Alexandr Savinov
2022-04-02 11:50:07 +02:00
simulated_trade_performance(df, 'sell_signal_column', 'buy_signal_column', 'close')
Alexandr Savinov
2022-03-20 10:09:33 +01:00
# Remove some items. Remove lists of transactions which are not needed
long_performance.pop('transactions', None)
short_performance.pop('transactions', None)
# Add some metrics. Add per month metrics
performance["profit_percent_per_month"] = performance["profit_percent"] / months_in_simulation
performance["transaction_no_per_month"] = performance["transaction_no"] / months_in_simulation
Alexandr Savinov
2022-08-20 11:21:25 +02:00
performance["profit_percent_per_transaction"] = performance["profit_percent"] / performance["transaction_no"] if performance["transaction_no"] else 0.0
performance["profit_per_month"] = performance["profit"] / months_in_simulation
long_performance["profit_percent_per_month"] = long_performance["profit_percent"] / months_in_simulation
short_performance["profit_percent_per_month"] = short_performance["profit_percent"] / months_in_simulation
Alexandr Savinov
2022-03-27 19:22:02 +02:00
performances.append(dict(
model=signal_model,
performance={k: performance[k] for k in ['profit_percent_per_month', 'profitable', 'profit_percent_per_transaction', 'transaction_no_per_month']},
long_performance={k: long_performance[k] for k in ['profit_percent_per_month', 'profitable']},
short_performance={k: short_performance[k] for k in ['profit_percent_per_month', 'profitable']}
Alexandr Savinov
2022-03-27 19:22:02 +02:00
))
Alexandr Savinov
2022-03-20 10:09:33 +01:00
#
Alexandr Savinov
2022-03-27 19:22:02 +02:00
# Flatten
Alexandr Savinov
2022-03-20 10:09:33 +01:00
#
Alexandr Savinov
2022-03-27 19:22:02 +02:00
# Sort
performances = sorted(performances, key=lambda x: x['performance']['profit_percent_per_month'], reverse=True)
Alexandr Savinov
2022-03-27 19:22:02 +02:00
performances = performances[:P.topn_to_store]
# Column names (from one record)
keys = list(performances[0]['model'].keys()) + \
list(performances[0]['performance'].keys()) + \
list(performances[0]['long_performance'].keys()) + \
list(performances[0]['short_performance'].keys())
Alexandr Savinov
2022-03-20 10:09:33 +01:00
lines = []
Alexandr Savinov
2022-03-27 19:22:02 +02:00
for p in performances:
record = list(p['model'].values()) + \
list(p['performance'].values()) + \
list(p['long_performance'].values()) + \
list(p['short_performance'].values())
record = [f"{v:.3f}" if isinstance(v, float) else str(v) for v in record]
Alexandr Savinov
2022-03-27 19:22:02 +02:00
record_str = ",".join(record)
lines.append(record_str)
Alexandr Savinov
2022-03-20 10:09:33 +01:00
#
# Store simulation parameters and performance
#
out_path = (out_path / App.config.get("signal_models_file_name")).with_suffix(".txt").resolve()
Alexandr Savinov
2022-03-20 10:09:33 +01:00
Alexandr Savinov
2022-03-27 19:22:02 +02:00
if out_path.is_file():
Alexandr Savinov
2022-03-20 10:09:33 +01:00
add_header = False
else:
add_header = True
Alexandr Savinov
2022-03-27 19:22:02 +02:00
with open(out_path, "a+") as f:
Alexandr Savinov
2022-03-20 10:09:33 +01:00
if add_header:
Alexandr Savinov
2022-03-27 19:22:02 +02:00
f.write(",".join(keys) + "\n")
Alexandr Savinov
2022-03-20 10:09:33 +01:00
#f.writelines(lines)
Alexandr Savinov
2022-07-17 21:32:45 +02:00
f.write("\n".join(lines) + "\n\n")
Alexandr Savinov
2022-03-20 10:09:33 +01:00
Alexandr Savinov
2022-03-27 19:22:02 +02:00
print(f"Simulation results stored in: {out_path}. Lines: {len(lines)}.")
Alexandr Savinov
2022-07-17 21:32:45 +02:00
elapsed = datetime.now() - now
print(f"Finished simulation in {str(elapsed).split('.')[0]}")
Alexandr Savinov
2022-03-20 10:09:33 +01:00
if __name__ == '__main__':
main()